Faraday Future is trying to reposition itself from a struggling electric-vehicle maker into a broader Embodied AI robotics company, with management now pointing to robotics sales, positive gross margins and U.S. manufacturing as the next phase of its turnaround. In its latest business update, the company said its robotics operation generated more than 30% average gross margin in the first half of 2026 and is targeting 2,000 cumulative robot sales and shipments by year-end.
Faraday Future’s latest strategy is becoming less about selling electric cars and more about putting artificial intelligence into machines that operate in the physical world.
The California-based company, which trades on Nasdaq under FFAI, says its EAI — or Embodied AI — robotics business has entered a new stage of commercial expansion, with sales scaling at positive gross margins while the company’s legacy automotive operation continues to weigh on consolidated financial performance.
In its latest weekly business update, Founder and Global CEO YT Jia said Faraday Future is pursuing a series of initiatives intended to turn robotics into a larger contributor to the company’s revenue and strategic value.
The headline target is ambitious: 2,000 cumulative robot sales and shipments by the end of 2026.
The company plans to concentrate its initial expansion in California, Texas, New York and the wider U.S. East Coast.
That commercial push sits within what Faraday Future calls its “Four-Core Full-Stack AI” strategy.
The first component is the physical robot itself, or what the company calls the EAI Robot Body. The second is the EAI Brain, covering the AI and robotics software stack. The third combines industry productivity solutions with a developer platform. The fourth is an EAI Data Factory, intended to generate the real-world training data needed to improve the company’s systems.
The strategy reflects a broader shift in robotics.
The industry’s competitive advantage is increasingly moving beyond mechanical hardware toward the combination of sensors, compute, foundation models, control systems and real-world data. Companies such as NVIDIA are building the infrastructure for that ecosystem, while robotics developers are attempting to use increasingly capable AI models to make machines more adaptable.
Faraday Future says it is deepening its integration with NVIDIA’s technology stack and advancing GR00T training and validation for complex grasping and multi-step operations.
The company also says it is moving its SONIC system from simulation toward whole-body control on physical robots.
The significance is that physical-world AI cannot be evaluated solely in software.
A model may perform well in simulation but still encounter problems when confronted with unpredictable objects, surfaces, lighting, human movement or mechanical limitations. The transition from simulated environments to real robots is therefore one of the central technical challenges in embodied AI.
Faraday Future’s planned EAI Data Factory is intended to address part of that problem.
The company says it wants to reach monthly capacity of 2,100 hours of qualified real-world data by the end of August and 20,000 hours by December, with a cumulative target of 50,000 hours during 2026.
If achieved, that data-generation pipeline could become an important component of its strategy: deploy robots, collect real-world interactions, use those experiences to improve AI models, and then feed improved capabilities back into additional robot deployments.
That creates a potential self-reinforcing development cycle.
But the company’s financial picture remains complicated.
Faraday Future says its robotics business generated an average gross margin above 30% during the first half of the year. Yet those economics have not translated into comparable consolidated profitability because the company’s automotive business continues to carry depreciation, engineering costs, fixed operating expenses and debt-related expenses.
This is perhaps the most important part of the update for investors.
A profitable robotics operation does not automatically resolve the financial obligations of the wider company.
Faraday Future says it reduced debt by more than $100 million year over year and plans to pursue a broader debt-reduction and resolution program during the second half of 2026.
Management is also exploring independent financing and a potential separate listing for the robotics business.
A separate listing would potentially allow investors to value the robotics operation independently from the legacy automotive business. However, it remains an exploration rather than a completed transaction, and the company has not provided enough detail to determine whether such a structure will ultimately proceed.
The manufacturing strategy is also changing.
Faraday Future says it is accelerating a “Built in USA” program covering seven FCC-certified robot models across three product series. The second phase is expected to focus on “Assembled in USA” production.
For an emerging robotics company, domestic assembly could serve several purposes beyond branding. It can help with supply-chain control, product customization and regulatory requirements while potentially making the company more attractive to U.S. industrial customers.
Faraday Future plans to discuss its manufacturing roadmap and recruit upstream and downstream partners through conferences scheduled for August 26 and September 28.
The company’s ambitions place it in a rapidly expanding but highly competitive robotics market.
NVIDIA provides much of the underlying accelerated-computing ecosystem. Companies including Tesla, Figure AI, Apptronik and others are pursuing different approaches to humanoid and general-purpose robotics. Industrial automation incumbents meanwhile continue to dominate established factory applications.
Faraday Future’s argument is that it can differentiate through a vertically integrated system combining robot hardware, AI, applications, developers and proprietary real-world data.
Whether that produces a durable advantage will depend on execution.
The company needs to demonstrate that its robots can perform useful commercial tasks reliably, that customers will buy them at profitable economics and that the data generated from deployments improves subsequent products.
Its stated target of 2,000 cumulative sales and shipments will therefore be an important test.
So will the company’s ability to separate improving robotics economics from the financial drag of its automotive legacy.
For now, Faraday Future is attempting something more complicated than an EV turnaround. It is trying to build a new identity around physical AI while using robotics revenue to create a path out of its legacy financial structure.
The opportunity is substantial. But the gap between an AI robotics demonstration and a scalable robotics business remains one of the industry’s hardest problems.
Market Landscape
The Embodied AI market is moving toward systems that combine AI models with physical machines, sensors, simulation and real-world data.
Key competitive layers include:
- Robot hardware: Humanoid and specialized robotic platforms.
- AI brains: Vision-language-action models and other systems that translate perception into physical actions.
- Accelerated computing: NVIDIA GPUs and robotics-focused compute platforms.
- Simulation: Virtual environments for training and validating robots before physical deployment.
- Real-world data: Demonstration and interaction data used to improve robotic policies.
- Developer ecosystems: APIs, SDKs and reusable skills that expand robot capabilities.
- Industrial applications: Manufacturing, logistics, inspection, education, security and other commercial deployments.
Faraday Future’s full-stack strategy is therefore competing not only with robotics manufacturers but with a broader ecosystem spanning NVIDIA, Tesla, Google, Microsoft, Amazon and specialist robotics startups.
The crucial enterprise question is shifting from whether a robot can perform a task to whether it can perform that task reliably and economically at scale.
Top Insights
- Faraday Future is shifting strategic emphasis toward Embodied AI robotics, reporting more than 30% first-half gross margins while targeting 2,000 cumulative robot shipments.
- Its four-core strategy combines robot hardware, an AI brain, developer solutions and real-world data infrastructure to build a vertically integrated robotics ecosystem.
- NVIDIA technology, GR00T training and SONIC whole-body control are central to FF’s effort to move AI capabilities from simulation into physical robots.
- The company is pursuing U.S. assembly across seven FCC-certified models, potentially strengthening domestic manufacturing, supply-chain control and regulatory positioning.
- Robotics profitability remains offset by automotive costs and legacy debt, prompting FF to explore financing options and a potential separate robotics listing.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
